2 replies to this topic

[Alfreedo]

 PSV calc 7-1PSV-142 Thermal To C2 Drum CheResources.pdf   12.98KB   57 downloadsFellow Engineers,

 

I am sizing a PSV for a thermal relief on the tube side of a Bitumen Vacuum Tower Bottoms(shell side)/Diluted Bitumen (tube side) exchanger.  The Exchanger datasheet lists the thermal transfer for clean service at 101 Btu/hr-ft^2-F.  There are 2 shells at 10,000 ft^2 each protected by one PSV.  The LMTD is 67 F.  The SG is about .83.  The Industry standard guidelines would put the cubic expansion coefficient at .0004.  When I calculate the same using hysis data, I get something larger at 0.0005.  If I plug all of these numbers in (including the .0005 expansion coefficient), I come up with a heat load of 135,000 Btu/hr and a required relief rate of 329 USGPM.  It sounds completely ridiculous to me for a thermal relief valve since normally TRVs are handled with a 1/2 x 3/4 relief valve. 

 

Two questions:

1)  Should I trust the expansion coefficient I calculated using HYSIS data or should I rely on the table from API 520 or 521 (not sure where it actually comes from).  Just doing that cuts the required volume to about 250 USGPM.

2)  Should I use the clean thermal transfer value if there is actually no flow on the tube side.  Certainly a stagnant fluid on one side of the exchanger must reduce the heat transfer efficiency.  If so, are there any rules of thumb to use by how much to reduce it. 

 

Here's the main reason I believe I must be doing something wrong.  The PSV has been in service for a long time and the initial calculated required relief rate was listed as 14 USGPM.  I can't find the calc so I don't know what they did to get that number.  Subsequently, another company did some flare study work for us and provided a number of 70 USGPM.  I am not privy to the calculation method.

 

Thanks for any input.

 

Best Regards,

 

Fred

Edited by Alfreedo, 24 September 2014 - 01:13 AM.

[Kwaku]

Alfredo, your relief rate seems very high. I would use the heat exchanger design duty if this information is available. I tried to calculate the LMTD based on the provided data on your sheet:

inlet temp (hot side)=246C

outlet temp (hot side)=191C

inlet temp (cold side)=170C

outlet temp (cold side)=204C

I had 30C (86F) instead of 67F. This could be corrected using the F factor. Just for clarification, the duty is around 135,000,000Btu/hr (not 135,000Btu/hr, but this was correct on your datasheet). Also, you did not provide the Specific Heat © of the cold fluid.

The formula I know for thermal expansion relief calculation is:

Q = (B*H)/(500*G*C)   API 521
where Q= Thermal relief rate (gpm), B = cubical expansion coefficient (1/F), H= heat transfer rate (btu/hr), G= specific gravity of the fluid and C= specific heat of the cold fluid (btu/lb.F). Hope this helps.

[Alfreedo]

 PSV calc 7-1PSV-142 Thermal To C2 Drum CheResources_2.pdf   13.08KB   28 downloadsThank you for your respsonse kwaku,

 

I believe since its a temperature difference, 30C becomes 30 C*1.8 = 54 F.  Yes.  I got my number from a HYSIS model which models a specific operational scenario that differes a bit from the datasheet.  In any case, using 54 brings my required rate down to 265.2 USGPM.  If I use what's shown as the required heat transfer rate (47.9 Btu/hr-ft^2-F) and drop the cubic expansion coefficient to .4 then my relief rate drops to 96.7 USGPM.  I'm just not sure that I can use that "required" value vs the "clean" value.  I will give the manufacturer a call to see if they can enlighten me on these numbers.

 

Fred

Edited by Alfreedo, 24 September 2014 - 12:24 PM.